Download PDF File

cm7209.qxd
04/12/2000
01:18
Page 85
Research Paper
85
The role of the cleavage site 2′-hydroxyl in the Tetrahymena
group I ribozyme reaction
Aiichiro Yoshida1*, Shu-ou Shan2*, Daniel Herschlag2 and Joseph A Piccirilli1
Background: The 2′-hydroxyl of U preceding the cleavage site, U(–1), in the
Tetrahymena ribozyme reaction contributes 103-fold to catalysis relative to a
2′-hydrogen atom. Previously proposed models for the catalytic role of this
2′-OH involve coordination of a catalytic metal ion and hydrogen-bond donation
to the 3′-bridging oxygen. An additional model, hydrogen-bond donation by the
2′-OH to a nonbridging reactive phosphoryl oxygen, is also consistent with
previous results. We have tested these models using atomic-level substrate
modifications and kinetic and thermodynamic analyses.
Results: Replacing the 2′-OH with -NH+3 increases the reaction rate ~60-fold,
despite the absence of lone-pair electrons on the 2′-NH+3 group to coordinate a
metal ion. Binding and reaction of a modified oligonucleotide substrate with
2′-NH2 at U(–1) are unaffected by soft-metal ions. These results suggest that
the 2′-OH of U(–1) does not interact with a metal ion. The contribution of the
2′-moiety of U(–1) is unperturbed by thio substitution at either of the
nonbridging oxygens of the reactive phosphoryl group, providing no indication
of a hydrogen bond between the 2′-OH and the nonbridging phosphoryl
oxygens. In contrast, the 103-fold catalytic advantage of 2′-OH relative to 2′-H
is eliminated when the 3′-bridging oxygen is replaced by sulfur. As sulfur is a
weaker hydrogen-bond acceptor than oxygen, this effect suggests a hydrogenbonding interaction between the 2′-OH and the 3′-bridging oxygen.
Conclusions: These results provide the first experimental support for the model
in which the 2′-OH of U(–1) donates a hydrogen bond to the neighboring
3′-bridging oxygen, thereby stabilizing the developing negative charge on the
3′-oxygen in the transition state.
Introduction
The Tetrahymena ribozyme (E) derived from a self-splicing
group I intron catalyzes a transesterification reaction:
CCC U CU p A + GOH → CCC U CU OH + Gp A
(S)
(P)
(1)
in which an exogenous guanosine nucleophile (G) cleaves
a specific phosphodiester bond of an oligonucleotide substrate (S; Table 1; see [1] and references therein). The
2′-OH of U preceding the scissile bond [U(–1)] facilitates
the chemical step ~103-fold relative to 2′-H, but does not
affect binding of S [2]. This corresponds to a contribution
of ~ 4 kcal/mol specifically to transition-state stabilization.
Linear free energy analysis with oligonucleotide substrates bearing a series of 2′-substituents at U(–1) suggested that the rate effect of the 2′-OH is more than that
expected from simple inductive effects [2]. Specific interactions of the 2′-OH may therefore contribute to catalysis.
How does this 2′-OH facilitate the reaction? Three models
can be postulated that are consistent with results from
Addresses: 1Departments of Biochemistry and
Molecular Biology, and Chemistry, University of
Chicago, 5841 S. Maryland Avenue, MC1028,
Chicago, IL 60637, USA. 2Department of
Biochemistry, B400 Beckman Center, Stanford
University, Stanford, CA 94305-5307, USA.
Correspondence: Daniel Herschlag;
Joseph A Piccirilli
E-mail: [email protected];
[email protected]
*These authors contributed equally to this work.
Key words: chemical modification, mechanistic
analysis, RNA catalysis
Received: 16 August 1999
Revisions requested: 22 September 1999
Revisions received: 7 October 1999
Accepted: 18 October 1999
Published: 11 January 2000
Chemistry & Biology 2000, 7:85–96
1074-5521/00/$ – see front matter
© 2000 Elsevier Science Ltd. All rights reserved.
previous biochemical studies and reasonable from geometric and chemical considerations: first, the 2′-OH interacts with an active-site metal ion that is crucial for
catalysis (Figure 1a); second, the 2′-OH donates a hydrogen bond to the 3′-bridging oxygen (Figure 1b); and
third, the 2′-OH donates a hydrogen bond to one of the
nonbridging oxygens of the reactive phosphoryl group
(Figure 1c). The first two models have been much discussed (Figure 1a,b; see below) [2–7], whereas the interaction with the reactive phosphoryl oxygen (Figure 1c)
has not been considered previously. Nevertheless, the
interactions postulated in these models have yet to be
tested directly.
A metal-ion interaction with the 2′-OH (Figure 1a) was
initially proposed on the basis of the observation that reaction of a CU (or CT) dinucleotide with a circular form of
the Tetrahymena intron has different Mg2+ concentration
dependencies with 2′-OH and 2′-H at U(–1) [3,8]. Similar
observations and proposals have been made with the RNA
component of RNase P [9]. In addition, the 2′-OH of the
guanosine nucleophile interacts with an active-site metal
cm7209.qxd
04/12/2000
86
01:18
Page 86
Chemistry & Biology 2000, Vol 7 No 2
Table 1
List of oligonucleotide substrates.
Abbreviation
I
II
III
IV
V
VI
VII
VIII
IX
X
XI
rSOH
rSH
rSN
dSOH
dSH
dSN
rSOH,3′–S
rSH,3′–S
rSOH,P–S
rSH,P–S
rSN,P–S
Oligonucleotide substrate
rC
rC
rC
dC
dC
dC
mC
mC
mC
mC
rC
–5
rC
rC
rC
dC
dC
dC
mC
mC
mC
mC
rC
rC
rC
rC
dC
dC
dC
mC
mC
mC
mC
rC
–3
rU
rU
rU
dU
dU
dU
rU
rU
rU
rU
rU
rC
rC
rC
dC
dC
dC
rC
rC
rC
rC
rC
–1
+1
rU-OpO-rA
dT-OpO-rA
nU-OpO-rA
rU-OpO-dA
dT-OpO-dA
nU-OpO-dA
rU-SpO-rA
dU-SpO-rA
rU-OpS-rA
dU-OpS-rA
nU-OpS-rA
rA
rA
+3
rA
rA
rA
rA
+5
rA
rA
dA
dA
dA
dA
dA
dA
dA
dA
dA
dA
dA
dA
S refers to the oligonucleotide substrate CCCUCUA5 or CCCUCUA,
without specification of the sugar identity. The presence of additional A
residues at the (+2) to (+5) position of S does not affect binding or the
reaction rate (R. Russell & D.H., unpublished observations). r refers to
2′-OH; d refers to 2′-H; m refers to 2′-OCH3; n refers to 2′-amino,
without specification of the protonation state. In the text, rSNH+3 and
rSNH2 refer to the oligonucleotide rSN with the 2′-amino group
protonated and deprotonated, respectively. 3'-S refers to the presence
of a thio substitution at the 3′-bridging oxygen of U(–1); P-S refers to
the presence of a thio substitution at the pro-SP oxygen of the reactive
phosphoryl group. 2′-OCH3 groups were introduced into the (–4) to
(–6) residues of 3′-S or P-S containing substrates; the sole effect of
this modification is to prevent miscleavage of these substrates, allowing
more accurate determination of reaction rates [28,59]. The chemical
composition of the reactive phosphoryl group is specified as follows:
ion (Figure 1, MC) [7,10], rendering an analogous metalion interaction with the 2′-OH of U(–1) an appealing
model. There are multiple metal ions within the active
site of the Tetrahymena ribozyme (Figure 1, MA, MB and
MC) [7,11–14]. The 2′-OH could, with reasonable geometry, coordinate the metal ion interacting with the 3′-bridging oxygen (Figure 1a, MA) [11], a metal ion interacting
with the pro-SP oxygen of the reactive phosphoryl group
[14], or an unidentified active-site metal ion. The previous
experimental results [3,8] do not distinguish, however,
between direct coordination of a metal ion by this 2′-OH
or indirect contribution of this 2′-OH to Mg2+ binding
through intervening active-site groups, nor is it known
whether the same reaction steps were followed with the
substrates bearing a 2′-OH and 2′-H.
A hydrogen bond between the 2′-OH and the 3′-bridging
oxygen (Figure 1b) was subsequently proposed [2]. The
Figure 1
Models for the catalytic role of the 2′-OH of
U(–1) within the Tetrahymena group I
ribozyme active site. The large bold letters
represent nucleotide bases on the
oligonucleotide substrate and the guanosine
nucleophile, the dashed lines (---) depict the
partial bond from the reactive phosphorus to
the 3′-OH of G and the 3′-oxygen of the
U(–1) residue of S, and ‘δ-’ depicts the partial
negative charges on the 3′-oxygens of U(–1),
the 3′-oxygen of G, and the nonbridging
oxygens of the reactive phosphoryl group in
the reaction’s transition state. MA is the metal
ion interacting with the 3′-bridging oxygen of
U (–1) [11], MB is the metal ion interacting
with the 3′-moiety of G [12], and MC is the
metal ion interacting with the 2′-moiety of G
[7,10]. (a) The 2′-OH of U(–1) interacts with
a catalytic metal ion. An interaction with MA is
depicted as an example of potential metal ion
interaction(s) with this 2′-OH; an alternative
interaction with another active-site metal ion is
also possible. (b) The 2′-OH donates a
hydrogen bond to the neighboring 3′-bridging
oxygen. (c) The 2′-OH donates a hydrogen
bond to one of the nonbridging oxygens of the
reactive phosphoryl group.
cm7209.qxd
04/12/2000
01:18
Page 87
Research Paper Cleavage site 2′′-OH in RNA catalysis Yoshida et al.
rate of the chemical step for the substrate with 2′-OH is
~ tenfold faster than that for the substrate with a 2′-fluoro
group at U(–1), despite the weaker electron-withdrawing
ability of 2′-OH than 2′-F [2]. As a 2′-fluoro group contains
lone-pair electrons that can accept hydrogen bonds but
cannot donate hydrogen bonds, the higher reactivity of the
substrate with 2′-OH than 2′-F at U(–1) suggests that
hydrogen-bond donation from the 2′-OH of U(–1) might
be important. A hydrogen bond from this 2′-OH to the
neighboring 3′-oxygen is the simplest model that accounts
for the specific transition-state stabilization provided by
the 2′-OH, as negative charge develops on the 3′-bridging
oxygen in the course of the reaction and could be stabilized by such a hydrogen bond [2,15–17]. This model has
been widely accepted, and extensions of this model have
been proposed (see below). Nevertheless, the model has
not been tested. Even if the 2′-OH of U(–1) acts as a
hydrogen-bond donor, the hydrogen bond could alternatively be donated to one of the nonbridging oxygens of the
reactive phosphoryl group, which may develop additional
negative charge in the transition state (Figure 1c). To our
knowledge, this interaction has not been discussed previously but is reasonable from geometrical considerations
and is consistent with all of the previous data.
Besides the interactions depicted in the models shown in
Figure 1, additional active-site interactions with the 2′-OH
of U(–1) have been suggested. The rate advantage provided by this 2′-OH relative to 2′-H is reduced ~fivefold
when the G•U wobble pair is replaced by a G•C pair [4],
and ≥ 30-fold when the 2′-OH of A207 is replaced by a
2′-H or 2′-F [6]. A crystal structure of an RNA duplex containing a G•U pair revealed an ordered water molecule
bridging the 2′-OH of U and the exocyclic amine of G
[18,19]. The biochemical data, in conjunction with the
crystallographic observations, led to the proposal of a
hydrogen-bonding network in which a hydroxyl group,
later identified as the 2′-OH of A207, donates a hydrogen
bond to the 2′-OH of U(–1) and accepts a hydrogen bond
from the exocyclic amine of G22, thereby bridging the
cleavage site 2′-OH and the G•U wobble pair [4,6,18].
Building on the earlier proposal [2], this network was suggested to orient the 2′-OH of U(–1) to favor its interaction
with the 3′-bridging oxygen (Figure 1b) [4,6]. Nevertheless, the results of these studies are also consistent with
the other models: the 2′-OH could alternatively be aligned
to interact with one of the nonbridging phosphoryl
oxygens (Figure 1c), or one of the lone-pair electrons of
this 2′-OH could accept a hydrogen bond from A207,
while the other electron pair interacts with an active-site
metal ion (Figure 1a).
In this work, each of the catalytic models depicted in
Figure 1 has been tested by atomic-level modifications of
the substrate combined with quantitative kinetic and
thermodynamic analysis. The results provide no indication
87
of a metal-ion interaction with the 2′-OH or a hydrogen
bond between this 2′-OH and a nonbridging reactive phosphoryl oxygen. In contrast, the results support the model
of Figure 1b, in which the 2′-OH of U(–1) donates a hydrogen bond to the 3′-bridging oxygen.
Results and discussion
We first describe experiments that test for metal-ion
interactions with the 2′-OH of U(–1) (Figure 1a). Two
independent approaches were used: replacing the 2′-OH
with 2′-NH3+ and investigating its effect on reactivity;
and replacing the 2′-OH with 2′-NH2 and exploiting the
preference of the -NH2 group to interact with metal ions
softer than Mg2+. The next section describes experiments that test hydrogen-bonding interactions of the
2′-OH with the nonbridging reactive phosphoryl oxygens
and with the 3′-bridging oxygen (Figure 1b and c,
respectively). These experiments support a hydrogen
bond between the 2′-OH and the 3′-bridging oxygen
(Figure 1b) and suggest that the other proposed interactions are not made (Figure 1a,c).
A metal-ion interaction with the 2′-OH of U(–1)?
To test whether a metal-ion interaction with the 2′-OH of
U(–1) is important for catalysis (Figure 1a), we investigated
the reactivity of the oligonucleotide substrate rSNH+3 , in
which the 2′-OH is replaced by -NH+3 (Table 1). The –NH+3
group has no lone pair electrons and therefore cannot interact with a metal ion. The rSNH+3 reaction would therefore
be expected to be severely compromised if the 2′-OH coordinates a metal ion important for catalysis (e.g. [10]).
To determine the reactivity of rSNH+3 , it was necessary to
establish the reaction conditions under which the 2′-amino
group of rSN (Table 1) is protonated and under which the
chemical step is rate limiting. We therefore first determined the effect of pH on the reaction of rSN and probed
whether the chemical step is rate limiting, as described
below. The pH dependence of the rSN reaction could
then be compared with those of rSOH and rSH to determine the effect of the 2′-NH+3 substitution on reactivity.
To follow the reaction of protonated rSNH+3, a pH value
below the pKa of the 2′-amino group must be used. (A
2′-NH+3 group deprotonates with a pKa of ~6.2 in aqueous
solution [20]; the pKa of this group is even lower with rSN
bound to the ribozyme, as described below.) Preliminary
experiments indicated that ribozyme activity is severely
compromised below pH 5 with the standard Mg2+ concentration of 10 mM, possibly due to disruption of ribozyme
structure at low pH ([21]; D.S. Knitt and D.H., unpublished
results). This deleterious effect of low pH can be overcome
at Mg2+ concentrations above 50 mM, as expected from stabilization of the active ribozyme structure by Mg2+ [22–24].
A Mg2+ concentration of 100 mM was therefore used in the
following experiments.
cm7209.qxd
04/12/2000
88
01:18
Page 88
Chemistry & Biology 2000, Vol 7 No 2
Table 2
Nomenclature for rate constants.
Rate constant
kG
c
(kc / Km)G
(kc/Km)S
Reaction
E•S•G → products
E•S + G → products
E•G + S → products
The rate of the reaction E•rSN + G → products [(kc/Km)G;
Table 2] was determined as a function of pH (Figure 2a,
filled circles). (kc/Km)G for rSN increases with increasing
pH below pH 5.0 but begins to level off at higher pH. To
isolate the effect that arises specifically from deprotonation
of the –NH+3 group, the pH dependence for the reaction of
the wild-type substrate rSOH was determined in parallel. As
observed previously at 10 mM Mg2+ [25], (kc/Km)G for rSOH
increases log-linearly with pH below pH 7.1 but levels off
at higher pH (Figure 2a, open circles). Comparison of the
pH dependencies for the rSN and rSOH reactions is complicated, however, as the apparent leveling off of (kc/Km)G for
rSOH at high pH results from a change in rate-limiting step
from chemical cleavage to a conformational step ([25] and
M. Khosla and D.H., unpublished observations). Indeed,
(kc/Km)G for rSH, for which the chemical step is ~103-fold
slower than rSOH and is rate limiting over the entire pH
range [25,26], has a log-linear pH dependence up to pH 8.5
(Figure 2a, open squares). The apparent leveling off of
Figure 2
rSNH+3 is more reactive than rSOH. (a) The pH dependencies of the rate
constant of reaction E•S + G → products [(kc / Km)G] with rSN (●),
rSOH (●
●) and rSH ( ■ ), determined as described in the Materials and
methods section (100 mM Mg2+). (b) The pH dependence of the effect
of thio-substitution at the pro-RP oxygen of the reactive phosphoryl group
G
on (kc / Km)G for rSN [thio effect = (kc / Km)G
oxy / (kc / Km)thio ]. The thio
effect at each pH represents the average of at least two independent
experiments, in which observed rate constants for substrates with and
without thio substitution were determined side-by-side in triplicate
((a) and data not shown). The dotted line shows the predicted pH
dependence of the thio effect if the rate-limiting step for the rSN
reaction above pH 5.0 changes to a step with a thio effect of 1.
(c) The pH dependence of the reactivity of rSN relative to rSOH
G
[(kc / Km)G,rel = (kc / Km)GrSN / (kc / Km)G
rSOH]. Values of (kc / Km) for each
substrate were from part (a). The line is a fit of the data to equation 3,
derived from equation 2 in the Results and discussion section (see the
Materials and methods section), and gives (kc / Km)G,rel
rSNH+3 = 58 ± 5 and
pKaE•rSN = 5.0 ± 0.2. The relative reactivity of rSNH2, (kc / Km)G,rel
rSNH2, could
not be determined, because the pH dependence of (kc / Km)G,rel would
need to be extended above pH 7.1, whereas interpretation of
(kc / Km)G,rel values above pH 7.1 is complicated by the change in ratelimiting step for the rSN reaction suggested by the decreased thio effect
((b); see text). Values of (kc / Km)G,rel were therefore plotted only from
pH 4.4 to 7.1. The arrow on the data point at pH 7.1 depicts that the
observed (kc / Km)G,rel is an upper limit for the relative reactivity of rSN in
the chemical step. This is because the chemical step is rate limiting for
(kc / Km)G of rSN at pH 7.1, whereas there is a change in the rate-limiting
step for (kc / Km)G of rSOH at this pH, as described in the text.
(kc/Km)G for the rSN reaction above pH 5.0 could arise from
different reactivities of rSNH+3 and rSNH2, a change in ratelimiting step analogous to the rSOH reaction, or a combination of these effects.
To differentiate between these possibilities, we examined the effect of thio substitution at the pro-RP oxygen
of the reactive phosphoryl group on the observed reaction
rate. The sole effect of this substitution is to slow down
the rate of the chemical step 2–3-fold for reaction of this
ribozyme, similar to effects observed with model compounds [25,27]. The thio modification slows the rate of
cm7209.qxd
04/12/2000
01:18
Page 89
Research Paper Cleavage site 2′′-OH in RNA catalysis Yoshida et al.
the rSN reaction by a constant amount, 1.9 ± 0.3-fold,
from pH 4.4 to 6.6 (Figure 2b). If the apparent leveling
off of (kc/Km)G for the rSN reaction represented a change
in the rate-limiting step, then the thio effect would be
expected to decrease above pH 5.0 (Figure 2b, dashed
line), where (kc/Km)G for rSN levels off (Figure 2a). The
thio effect therefore suggests that the chemical step is
rate limiting for reaction of rSN above pH 5.0. Above
pH 7, the thio effect begins to decrease (Figure 2b), possibly reflecting a change in rate-limiting step for the rSN
reaction at the highest pH values, analogous to rSOH.
Reactions at these high pH values were therefore not
analyzed here.
As the chemical step is rate limiting for both the rSN and
rSOH reaction below pH 7.1, the effect on reactivity of rSN
that arises specifically from deprotonation of the –NH+3
group can be isolated by plotting the values of (kc/Km)G for
rSN relative to that for rSOH as a function of pH
G / (k / K ) G ; Figure 2c]. At the
[(kc / Km)G,rel = (kc/Km)rS
c
m rSOH
N
lowest pH values (4.4–4.7), rSN reacts 44-fold faster than
rSOH. (kc/Km)G,rel decreases with increasing pH above
pH 5.0 (Figure 2c), suggesting that rSNH2 is less reactive
than rSNH+3 :
(2)
A fit of the pH dependence of (kc/Km)G,rel in Figure 2c to
the model of equation 2 gives (kc/Km)G,rel
rSNH+3 = 58 ± 5 for the
•rSN = 5.0 ± 0.2
reactivity of rSNH+3 relative to rSOH, and pKE
a
for deprotonation of the 2′-NH+3 group in the E•rSN
complex. This pKa value is lower than the solution pKa
value of 6.2 for the 2′-amino group [20], presumably
because of perturbation of the pKa when rSN is bound to
the ribozyme (see the Materials and methods section and
[20,28] for details). As (kc/Km)G,rel continues to decrease
from pH 5.0 to 7.1, the reaction observed over this pH
range arises from a small fraction of the more reactive
species rSNH+3 , whereas rSNH2 makes a negligible contribution to the observed reaction rate.
The higher reactivity of rSNH+3 than rSOH, despite the
absence of lone-pair electrons on the 2′-NH+3 group to
interact with a metal ion, suggests that a metal-ion interaction with the 2′-OH of U(–1) is not important for catalysis by the Tetrahymena ribozyme. An alternative model, in
which a metal-ion interaction with the 2′-OH in the wildtype reaction is replaced by the positively charged –NH+3
group, remains possible but not likely, because replacing a
89
catalytic interaction with a fortuitous interaction would not
be expected to increase reactivity.
To probe independently for a metal ion coordinated by the
2′-moiety of U(–1), we examined the effect of soft-metal
ions on the reactivity of the oligonucleotide substrate with
a 2′-NH2 group at U(–1). These experiments are based on
the observation that nitrogen ligands typically prefer to
interact with soft-metal ions than with Mg2+, whereas
oxygen ligands do not exhibit such preferences [10,29,30].
This approach has previously been used to identify a
metal-ion interaction with the 2′-OH of the guanosine
nucleophile in this ribozyme (Figure 1, MC) [7,10]. Analogous metal-ion rescue of the reaction of modified oligonucleotide substrates with 2′-NH2 at U(–1) might be
expected if a metal-ion interaction with the 2′-moiety of G
is crucial for catalysis.
To provide a most sensitive probe for a potential metal-ion
interaction with the 2′-NH2 of U(–1), the reaction
E•G + S → products [(kc/Km)S; Table 2] was followed. In
this reaction, S is not bound to the ribozyme in the starting
ground state, so that any interactions with active-site metal
ions would be formed in the course of the reaction, allowing the observed reaction rate to provide a probe for such
an interaction. To ensure that the chemical step was rate
limiting, the oligonucleotide substrate dSN (Table 1) was
used. Replacing the 2′-OH groups of S with 2′-H at
residues other than U(–1) weakens binding of S, but does
not affect the rate of the chemical step once S is bound
within the ribozyme active site [31]. The pH dependence
of (kc/Km)S and additional observations strongly suggest
that the chemical step is rate limiting for the dSN reaction
and, furthermore, that the reaction of dSNH2 can be followed at pH 7.5, without significant contribution from reaction of the small fraction of dSNH+3 present at this pH (0.05;
pKa ~ 6.2 [20]; see the Materials and methods section).
The effect of soft-metal ions on the rate constant of the reaction E•G + dSNH2 → products was determined in a background of 10 mM Mg2+. Addition of Mn2+, up to 10 mM, had
the same effect on the reaction E•G + S → products with
dSNH2 as with dSH (Figure 3a,b; 10 mM Mg2+). Mn2+ does
not, therefore, provide a specific stimulation of the reaction of dSNH2 (relative to dSH). Similarly, addition of
several other soft-metal ions, Zn2+, Cd2+ and Co2+, did not
provide a specific stimulation for reaction of dSNH2 relative to dSH at concentrations up to 0.5, 0.5 and 5 mM,
respectively (data not shown).
To allow Mn2+ to compete more effectively with Mg2+ at
the potential metal site, we lowered the concentration of
Mg2+ to 0.1 mM and used even higher Mn2+ concentrations. Nevertheless, addition of up to 50 mM Mn2+ did
not provide a specific stimulation for reaction of dSNH2
relative to dSH [Figure 3c,d; analogous results were
cm7209.qxd
04/12/2000
90
01:18
Page 90
Chemistry & Biology 2000, Vol 7 No 2
Figure 3
Mn2+ does not stimulate the reaction of
dSNH2. (a) Effect of Mn2+ on the rate constant
of the reaction E•G + S → products
[(kc / Km)S] with dSN (●), dSOH (■) and dSH
(●
●). The second-order rate constant (kc / Km)S
was determined at pH 7.5 with 10 mM Mg2+
background, as described in the Materials and
methods section. The dashed line denotes
that (kc / Km)S for dSOH above 1 mM Mn2+ is
limited by the binding of the substrate instead
of the chemical step (S.S. and D.H.,
unpublished observations). (b) Effect of Mn2+
on the reactivity of dSNH2 relative to dSH in
the reaction: E•G + S → products
S / (k / K ) S ].
[(kc / Km)S,rel = (kc / Km)dS
c
m dSN
N
Values of (kc / Km)S were from part (a). The
straight line with a slope of zero denotes the
absence of an effect of Mn2+ on (kc / Km)S,rel.
(c) Same as in (a) except that a background
of 0.1 mM Mg2+ was used instead of 10 mM.
(d) Same as in (b) except that a 0.1 mM Mg2+
background was used. Values of (kc / Km)S
were from (c). The straight line with a slope of
zero denotes the absence of an effect of Mn2+
on (kc / Km)S,rel.
obtained in reactions with a background of 2 mM Mg2+
(data not shown)]. Thus, if a Mn2+ could interact with the
2′-NH2 group more strongly than Mg2+ relative to 2′-OH,
this Mn2+ would need to bind much more weakly than
Mg2+ to this metal-ion site in the E•G complex. For
example, if replacing the Mg2+ at the potential metal site
with Mn2+ provides a tenfold stimulation for the dSNH2
reaction relative to dSH, then this metal site would bind
Mg2+ 5000-fold more strongly than Mn2+ [Mg2+ specificity = ([Mn2+] / [Mg2+]) × rate enhancement = (50/0.1) ×
10 = 5000]. An even greater Mg2+ specificity would be
required if the rate enhancement provided by the potential Mn2+ ion were larger.
The absence of metal-ion rescue of the dSNH2 reaction is in
marked contrast to the ability of Mn2+ and Zn2+ to rescue
the reaction of GNH2, in which the 2′-OH of G is replaced
by -NH2, in the Tetrahymena ribozyme and other group I
ribozymes (20-fold and ≥ 60-fold rescue of the GNH2 reaction by Mn2+ and Zn2+, respectively) [7,10]. In solution,
compounds containing nitrogen ligands typically interact
10–102-fold more strongly with Mn2+ and 102–103-fold more
strongly with Zn2+, Co2+ and Cd2+ than with Mg2+, whereas
compounds containing exclusively oxygen ligands do not
typically exhibit such strong preferences [29,30]. Finally,
substantially stronger binding of Mg2+ than Mn2+ to the
potential metal-ion site is contrary to previous studies of
metal-ion–RNA interactions, in which similar or stronger
binding of Mn2+ than Mg2+ was observed (e.g. [10,32–34]).
Although a metal-ion interaction with the 2′-OH cannot be
ruled out on the basis of negative results, these results are
most simply accounted for by the absence of a metal-ion
interaction with the 2′-moiety of U(–1).
In summary, the higher reactivity of the substrate with a
2′-NH+3 group than a 2′-OH group at U(–1) and the
absence of metal-ion rescue of reaction of the substrate
with 2′-NH2 suggest that the 2′-OH of U(–1) does not
interact with an active-site metal ion. This and previous
results [13] suggest an asymmetry in transition-state interactions at the Tetrahymena ribozyme active site, with a
metal ion interacting with the 2′-OH of the guanosine
nucleophile but not with the 2′-OH of U(–1). The absence
of a metal-ion interaction with the 2′-OH of U(–1) also
cm7209.qxd
04/12/2000
01:18
Page 91
Research Paper Cleavage site 2′′-OH in RNA catalysis Yoshida et al.
provides an explanation for the substantially larger deleterious effect of removing the 2′-OH of the guanosine nucleophile than the 2′-OH of U(–1) (> 106-fold compared with
103-fold reduction in the rate of the chemical step; [35,36],
and S. S. and D.H., unpublished observations), as placing
a hydrogen atom adjacent to a metal ion is likely to have
major energetic consequences.
Hydrogen-bond donation from the 2′′-OH of U(–1)?
The faster chemical step for oligonucleotide substrates
with 2′-OH than with 2′-F at U(–1) suggested that hydrogen-bond donation from the 2′-OH of U(–1) contributes
to catalysis (see above) [2]. As described in the Introduction section, however, there are three substrate atoms
that could potentially accept a hydrogen bond from this
2′-OH: the neighboring 3′-bridging oxygen, or the pro-RP
or pro-SP oxygen of the reactive phosphoryl group
(Figure 1b and c, respectively). In this section, each of
these potential hydrogen-bonding interactions was tested
by replacing the oxygen with sulfur. As sulfur is a weaker
hydrogen-bond acceptor than oxygen ([37–42] and references therein), thio substitution would be expected to
diminish the catalytic advantage provided by a hydrogen
bond from the 2′-OH.
91
Several previous observations suggest that there are no catalytically important active-site interactions with the pro-RP
oxygen of the reactive phosphoryl group. Thio substitution of the pro-RP oxygen reduces the rate of the chemical
step only 2–3-fold, consistent with the effects observed in
reactions of model compounds ([25,27]; see also above). In
addition, similar thio effects at the pro-RP oxygen were
observed with substrates containing 2′-OH or 2′-H at
U(–1) ([25] and data not shown), suggesting that there is
no interaction between the 2′-OH and the pro-RP oxygen.
Thio substitution at the pro-SP oxygen reduces the rate of
the chemical step ~103-fold ([27] and see below), strongly
suggesting the presence of active-site interactions with the
pro-SP oxygen (e.g. [14]). To probe a potential hydrogen
bond between the pro-SP oxygen and the 2′-OH of U(–1),
the effect of thio substitution at this oxygen was investigated with substrates bearing a series of 2′-substituents
(Table 1, substrates I–III compared with IX–XI). Reactions
were carried out under conditions such that the chemical
step is rate limiting for each substrate (Table 3). To facilitate accurate determination of rate constants, we used saturating ribozyme concentrations with respect to S, such that
the reactions E•S + G → products and E•S•G → products
Table 3
Thio-effects at the pro-Sp oxygen of the reactive phosphoryl group with different 2′′-substituents at U(–1).*
2′-X†
pH
OH
5.5
6.2
H
(kc / Km)G (M–1min–1)
Thio effect§
pH
–1
kG
c (min )
Thio effect§
P–O
1.3 × 104
1.1 × 105
P–S
2.0
17
6.5 × 103
6.5 × 103
4.4
4.4‡
5.0
5.0‡
P–O
0.13
0.29
0.45
1.05
P–S
6.5 × 10–5
1.9 × 10–4
4.6 × 10–4
9.1 × 10–4
2.0 × 103
1.5 × 103
0.98 × 103
1.2 × 103
—
—
—
7.9
7.5
7.0
0.25
0.16
0.034
2.3 × 10–4
9.7 × 10–5
2.6 × 10–5
1.1 × 103
1.6 × 103
1.3 × 103
6.7 × 103
6.4 × 103
–
–
—
—
–
—
NH3
+
4.4‡
4.7‡
1.6 × 104
7.7 × 104
2.4
12
NH2
7.5
2.3 × 105
64
3.6 × 103
*Rate constants (kc/Km)G and kG
c were determined with substrates
containing oxygen or sulfur at the pro-SP position of the reactive
phosphoryl group (P–O and P–S, respectively; Table 1, I–III and IX–XI)
in side-by-side experiments under conditions such that the chemical
step is rate limiting, as described in the Materials and methods section,
and were measured at 10 mM Mg2+ unless otherwise specified. Each
rate constant represents the average of at least two independent
determinations that vary by < 20%. Addition of up to 1 mM DTT or
EDTA had no significant effect on the observed rate constants
(< 20%). —, not determined. †2′-X refers to the 2′-substituent at U(–1).
‡Determined at 100 mM Mg2+. §The thio effect is 4–5-fold larger for
(kc/Km)G than for kG
c . This is presumably because the P–O and P–S
containing substrates bind the ribozyme in different conformations.
The ribozyme binds oligonucleotide substrates in two steps: formation
of an open complex, in which S is held solely by base-pairing
interactions with the internal guide sequence of E to form the P1
duplex, followed by docking of P1 into the ribozyme core via tertiary
interactions to form the closed complex ([59–61] and references
therein). The P–O-containing substrates bind the ribozyme to form
closed complexes, whereas the P–S-containing substrates bind the
ribozyme to form open complexes (data not shown). The guanosine
nucleophile binds ~fivefold more strongly to the closed complex than
to the open complex [58]. Thus, binding of G to the E•S complexes
formed by the P–O-containing substrates is expected to be ~fivefold
stronger than those formed by the P–S-containing substrates. This has
been verified for several substrates (data not shown).
cm7209.qxd
04/12/2000
92
01:18
Page 92
Chemistry & Biology 2000, Vol 7 No 2
were followed [(kc/Km)G and kGc , respectively; Table 2].
Thio substitution of the pro-SP oxygen (P–O → P–S) has a
similar effect on the rate constants of each substrate, regardless of whether a 2′-OH, 2′-H, 2′-NH+3 or 2′-NH2 is present
at U(–1) (Table 3). The observed thio effects are independent of pH and Mg2+ concentration (Table 3). The constant
thio effect observed with different 2′-substituents at U(–1)
suggests that the catalytic contribution of the 2′-substituent
at U(–1) is independent of the presence of sulfur or oxygen
at the pro-SP position of the reactive phosphoryl group.
There is therefore no indication of a direct hydrogen bond
between the 2′-OH and the pro-SP oxygen.
Finally, we probed the hydrogen bond between the
2′-OH and the 3′-oxygen of U(–1) (Figure 1b) by replacing this 3′-oxygen with sulfur. The rate of the reaction
E•S•G → products (kG
c ) was determined for 3′-thio substituted substrates with either 2′-OH or 2′-H at U(–1)
[rSOH,3′-S and rSH,3′-S, respectively (Table 1)]. In the presence of 10 mM Mg2+, the rate constant of the chemical
step for the thio-substituted substrates is similar with
2′-OH or 2′-H at the cleavage site (within twofold;
Table 4). The experiment was repeated with added Mn2+,
as the presence of 3′-bridging sulfur severely compromises reactivity in Mg2+ alone, whereas reaction of 3′-thio
substituted substrates can be rescued by replacing the
Mg2+ at site A with Mn2+ (Figure 1b, MA; [11] and data
not shown). Mn2+ binds to metal site A with an apparent
dissociation constant of KMn,app = 0.8 mM (10 mM Mg2+;
[13]), so that 10 mM Mn2+ allows Mn2+ to fully replace
Mg2+ at site A. With metal site A occupied by Mn2+, rate
constants for reaction of 3′-thio-substituted substrates are
similar with 2′-OH or 2′-H at U(–1) (Table 4). Thus, the
2′-OH at U(–1) does not make a significant contribution
to the chemical step with substrates containing a 3′-sulfur
leaving group, whether Mg2+ or Mn2+ is bound at metal
site A. This is in marked contrast to the 103-fold faster
reaction with 2′-OH than 2′-H for substrates with an
oxygen at the 3′-bridging position (Table 3) [2]. The
removal of the catalytic advantage of the 2′-OH of U(–1)
by the 3′-thio substitution is most simply interpreted in
terms of a direct hydrogen bond between the 2′-OH and
the 3′-bridging oxygen (Figure 1b).
In summary, the results presented in this section provide
no indication of a hydrogen bond between the 2′-OH of
U(–1) and the nonbridging oxygens of the reactive phosphoryl group (Figure 1c). In contrast, the results provide
experimental support for the catalytic model previously
proposed, in which the 2′-OH of U(–1) donates a hydrogen bond to one of the lone-pair electrons on the 3′-bridging oxygen in the transition state (Figures 1b and 4)
[2,4,6]. The 3′-leaving group oxygen is partially anionic in
the transition state, favoring formation of such a hydrogen
bond; in contrast, this oxygen is electron-deficient in the
ground state, rendering this hydrogen-bonding interaction
less favorable [2,43,44]. Donation of a hydrogen bond
from the neighboring 2′-OH to the 3′-bridging oxygen
presumably provides a specific stabilization of the transition state relative to the ground state, thereby facilitating
the chemical step. The developing negative charge on the
leaving oxygen atom is apparently further stabilized by a
metal ion, which interacts with the other lone pair of electrons on this oxygen (Figures 1b and 4, MA; [11]; A.Y.
et al., unpublished observations). It appears that the
Tetrahymena ribozyme focuses many of its active-site interactions on this 3′-leaving-group oxygen to provide rate
acceleration. This is sensible as this atom presumably
undergoes the largest amount of charge rearrangement in
the course of the reaction.
Alignment of the 2′′-OH by the ribozyme active site?
Although the 2′-OH of U(–1) appears to stabilize the developing negative charge on the 3′-bridging oxygen in the
Tetrahymena ribozyme reaction, the 2′-OH has only a
modest effect on the pKa value of the 3′-OH group in
aqueous solution (the pKa values of the 3′-OH of ribose and
2′-deoxyribose are 12.1 and 12.6, respectively [45,46]). This
Table 4
Contribution of the 2′-OH at U(–1), relative to 2′-H, on the rate of the chemical step (kGc ) with a 3′-sulfur or 3′-oxygen leaving group.*
MA†
3′-O
–1
kG
c (min )
Mg2+
Mn2+
2′-OH
0.84
–
2′-H
9.7 × 10–4
–
3′-S
krel ‡
870
–
*The rate constant kG
c was determined with oligonucleotide substrates
containing oxygen or sulfur at the 3′-leaving group (3′-O and 3′-S,
respectively; Table 1, I–II and VII–VIII) in side-by-side experiments under
conditions such that the chemical step is rate-limiting, as described in
the Materials and methods section. The rate constants reported for
substrates with 3′-O were determined at pH 5.2, and those with 3′-S
were determined at pH 7.0. Varying the pH from 5.0 to 8.1 did not
–1
kG
c (min )
2′-OH
2.5 × 10–5
7.0 × 10–2
2′-H
1.4 × 10–5
7.8 × 10–2
krel ‡
1.8
0.9
significantly change the krel values for the 3′-S containing substrate
(10 mM Mn2+ / 10 mM Mg2+; data not shown). †MA refers to the metal
ion (Mg2+ or Mn2+) bound at metal site A (Figure 1, MA). Values of kG
c
with Mg2+ bound at site A were determined at 10 mM Mg2+, and those
with Mn2+ bound were determined at 10 mM Mn2+ / 10 mM Mg2+ (see
G
text). ‡krel = (kG
c ,2′-OH)/(kc ,2′-H) is the rate constant for reaction of the
substrate with 2′-OH at U(–1) relative to that with 2′-H.
cm7209.qxd
04/12/2000
01:18
Page 93
Research Paper Cleavage site 2′′-OH in RNA catalysis Yoshida et al.
suggests that the 2′-OH does not significantly stabilize the
adjacent 3′-oxyanion in aqueous solution. Why, then, is the
ribozyme reaction 103-fold faster with a 2′-OH than a 2′-H
at U(–1)? As described in the Introduction section, the catalytic contribution of the 2′-OH of U(–1) is reduced upon
mutating the G•U wobble pair or upon modifying the
2′-OH of A207 [4,6]. These results, in conjunction with
crystallographic observations (see the Introduction section)
[18,19], suggest a network of interactions surrounding the
2′-OH of U(–1) within the ribozyme active site that is not
present in aqueous solution. As proposed previously [6],
the 2′-OH of A207 may donate a hydrogen bond to the
2′-OH of U(–1) and bridge between this 2′-OH and the
exocyclic amine of the G•U wobble pair (Figure 4).
Nevertheless, it is difficult to ascribe catalytic contributions to individual active-site groups; the difficulty is more
pronounced for the groups discussed here as they are interconnected to form a network. Two extreme models can be
presented that are both consistent with the previous data.
As suggested previously, interactions with A207 and the
G•U pair may be used to orient the 2′-hydroxyl proton of
U(–1), thereby favoring formation of the hydrogen bond to
the 3′-oxygen in the ribozyme reaction [4,6]. Alternatively,
the network of interactions surrounding the 2′-OH may
limit rearrangement within the active site relative to
aqueous solution, such that replacing the 2′-OH by 2′-H is
particularly deleterious for the ribozyme reaction. That is,
in aqueous solution, the hydroxyl groups from water can
stabilize the 3′-oxyanion when the 2′-OH is removed, such
that this modification has no significant effect on the pKa
value of the 3′-OH. In contrast, in the ribozyme active site,
access to water may be limited because the pocket created
by removal of the 2′-OH may not be sufficient to readily
accommodate a water molecule and because there would
be a substantial energetic penalty for rearranging the
active-site network surrounding the 2′-OH.
The network of interactions between the 2′-OH of U(–1),
the 2′-OH of A207 and the exocyclic amine of G22 have
been referred to as a ‘catalytic triad’ [6]. The energetic
behavior of this network appears to be distinct, however,
from the highly cooperative interactions found within the
catalytic triad of serine proteases, in which mutation of any
of the sidechains essentially removes the catalytic contribution of other groups within the triad [47]. In contrast,
although an energetic interaction between the 2′-OH of
U(–1) and the G•U wobble pair was suggested from the
fivefold reduction in the contribution of the 2′-OH upon
changing the G•U wobble to a G•C pair (Figure 4) [4], a
catalytic contribution of 102-fold from this 2′-OH (relative
to 2′-H) remains even with a G•C pair at the cleavage site.
Furthermore, after removal of the exocyclic amine of G22,
the effect of removing the 2′-OH of U(–1) remains within
twofold of that of the wild-type ribozyme [5], and replacing
the 2′-OH of A207 with 2′-H or 2′-F still has a deleterious
93
Figure 4
Model of transition state interactions within the Tetrahymena ribozyme
active site. The dashed lines (---) depict the partial bond from the
reactive phosphorus to the 3′-OH of G and the 3′-oxygen of the U(–1)
residue of S, and ‘δ-’ depict the partial negative charges on the 3′oxygens of U(–1) and G, as described in Figure 1. MA, MB and MC are
the three active-site metal ions that interact with the 3′-bridging oxygen
of U (–1) and the 3′- and 2′-OH of G, as described in the legend to
Figure 1 [7,10–12]. The results described here provide experimental
evidence for a hydrogen bond from the 2′-OH of U(–1) to the
3′-bridging oxygen, as was previously proposed [2]. Previous work
suggests that this 2′-OH may also participate in a hydrogen-bonding
network that involves the 2′-OH of A207 and the G•U wobble pair that
precedes the cleavage site [4,6,18].
effect on ribozyme function [6]. These observations
suggest that the exocyclic amine of G22 is not required for
the 2′-OH of U(–1) or the 2′-OH of A207 to make its interactions. In addition, these energetic effects suggest the
presence of a more extensive network of interactions
within this RNA active site that restricts positioning even
in the absence of the exocylic amine of G22. In contrast,
the observation that the chemical step for rSNH+3 is faster
than rSOH (Figure 2 and equation 2), despite the absence of
lone-pair electrons on the 2′-NH+3 group to accept a hydrogen bond from A207, suggests that there is some ability to
rearrange active-site groups within this network. It will be
fascinating to explore further the energetic behavior of this
active-site network within the Tetrahymena ribozyme, to
unravel the structural basis for its behavior, and to compare
its behavior with those of other RNA and protein enzymes.
Catalytic role of the cleavage site 2′′-OH in other RNA
enzymes
The 2′-OH at the cleavage site of group II introns makes
only a small contribution to catalysis; replacing this 2′-OH
by 2′-H reduces the observed reaction rate by less than
30-fold [48]. This suggests that the cleavage site 2′-OH is
not used by group II introns for catalysis. The ability of
group II introns to utilize DNA substrates could facilitate
their integration into new genomic positions and is presumably important for the mobility of some group II introns
cm7209.qxd
04/12/2000
94
01:18
Page 94
Chemistry & Biology 2000, Vol 7 No 2
(e.g. [49–52] and references therein). The small effect of
removing the cleavage site 2′-OH in group II introns contrasts with the 103-fold effect observed in the Tetrahymena
group I ribozyme, and provides further support for the
presence of a network of active-site interactions within the
Tetrahymena RNA active site that accentuates the energetic
cost for removal of this 2′-OH.
Large catalytic contributions of the 2′-OH preceding the
scissile bond have also been observed with several other
RNA enzymes that carry out reactions similar to that catalyzed by the Tetrahymena group I ribozyme. Replacing the
2′-OH by a 2′-H results in a ~2000-fold reduction in reaction
rate for the Anabaena group I ribozyme [53], and a 3400-fold
rate reduction for the RNA component of RNase P [9].
Mechanistic models such as stabilization of the neighboring
3′-leaving-group oxygen and stabilization of bound activesite metal ion(s) have been proposed (e.g. [9]) but remain to
be tested. Experiments analogous to those described here
might be useful in delineating the catalytic role of the cleavage site 2′-OH groups in these ribozyme reactions.
Significance
The discovery of RNA catalysis has generated much
biochemical and mechanistic interest. Delineating the
catalytic interactions utilized by an RNA enzyme and
dissecting the properties of its active site are central to
understanding how RNA enzymes are able to act as
efficient biological catalysts. In this work, site-specific
substrate modifications combined with quantitative
analysis have been used to test several catalytic models
proposed for the 2′-OH of U(–1) in the Tetrahymena
group I ribozyme reaction. The results provide experimental support for a previously proposed model in
which the 2′-OH donates a hydrogen bond to the neighboring 3′-bridging oxygen, thereby stabilizing the developing negative charge on the leaving group oxygen in the
transition state. These results demonstrate the utility of
chemical modification combined with in-depth mechanistic analysis as a tool for delineating the catalytic roles
of individual active-site groups. Analogous approaches
can be used to dissect catalytic interactions in the active
sites of other RNA and protein enzymes (e.g. [54]).
The large deleterious effect of removing the 2′-OH of
U(–1) suggests that this 2′-OH is situated within a
network of active-site interactions. A model has been
proposed in which the 2′-OH of U(–1), the 2′-OH of
A207 and the exocylic amine of G22 forms a hydrogenbonding network. Nevertheless, the energetic behavior
of this network is complex, suggesting that the contribution of the individual groups within this network cannot
be simply dissected. The presence of additional functional groups involved and the structural bases for the
energetic behavior of this network remain important
questions for future investigations.
Materials and methods
Materials
Ribozyme was prepared by in vitro transcription with T7 RNA polymerase
as described previously [55]. Oligonucleotides were made by solid-phase
synthesis and supplied by the Protein and Nucleic Acid Facility at Stanford University or were gifts from L. Beigelman (Ribozyme Pharmaceuticals Inc.) or F. Eckstein. Oligonucleotides containing a bridging sulfur at
the 3′-moiety of U(–1) were synthesized by published procedures [56].
Oligonucleotide substrates were 5′-end-labeled using [γ-32P]ATP and T4
polynucleotide kinase and purified by electrophoresis on 24% nondenaturing polyacrylamide gels, as described previously [31,55].
General kinetic methods
All reactions were single turnover, with ribozyme in excess of labeled
oligonucleotide substrate (S*) and were carried out at 30°C in 50 mM
buffer. The buffers used were: sodium acetate, pH 4.4–5.6; NaMES,
pH 5.4–7.0, NaMOPS, pH 6.4–7.1, NaHEPES, 6.8–7.5, NaEPPS,
pH 7.5–8.5 (pH values determined at 30°C). Reactions were followed
and analyzed essentially as described previously [57,58]. Briefly,
ribozyme was preincubated in 10 mM MgCl2 and 50 mM buffer at 50°C,
cooled to 30°C, and Mg2+, Mn2+, Zn2+, Cd2+ or Co2+ was added to
obtain the desired metal ion concentrations prior to initiating the reaction
by addition of S* (< 0.1 nM). For reactions with 0.1 mM Mg2+ background, ribozyme was folded in 2 mM MgCl2 at 50°C, cooled to 30°C,
and adjusted to desired metal ion concentrations. For reactions carried
out above pH 8.0, the preincubation was carried out at pH 7.5 to avoid
degradation and diluted tenfold into the appropriate buffer at 30°C [21].
Six aliquots of 2 µl reaction mixture were removed from 20 µl reactions at
specified times, and further reaction was quenched by addition of 4 µl of
stop solution [90% formamide with EDTA in ≥ twofold excess of total
divalent metal ion (20–200 mM), 0.005% xylene cyanole, 0.01% bromophenol blue and 1 mM Tris, pH 7.5]. Oligonucleotide substrates and
products were separated by electrophoresis on 20% polyacrylamide/7 M
urea gels, and their ratio at each time point was quantitated with a Molecular Dynamics Phosphorimager.
Reactions were followed for ≥ 3t1/2 except for very slow reactions.
Good first-order fits to the data, with endpoints of ≥ 90%, were
obtained (KaleidaGraph). The slow reactions were typically linear for up
to 20 h, and an endpoint of 95% was assumed to obtain observed rate
constants from the initial rates.
Kinetic constants
The nomenclature used for rate constants is defined in Table 2. kG
c , the
first-order rate constant, was determined for oligonucleotide substrates I,
II and VII–X (Table 1) with E saturating with respect to S (50–1000 nM
E; KSd ≤ 8 nM; data not shown) and with saturating G (2 mM; KG
d ≤
500 µM). The second-order rate constant (kc/Km)G was determined for
oligonucleotide substrates I–III and IX–XI (Table 1) with E saturating with
respect to S as above, but with G subsaturating (≥ three concentrations,
S
each > fivefold below KG
d ). The second-order rate constant (kc/Km) was
determined for oligonucleotide substrates IV–VI (Table 1) with saturating
G (2 mM) but with E subsaturating with respect to S (10–50 nM E; KSd ≥
150 nM; data not shown).
Following the chemical step for reaction of dSNH2
To determine the effect of metal ions on the reactivity of dSNH2, it is
crucial that the chemical step is rate limiting and that the reaction of
deprotonated dSNH2, rather than dSNH+3 , was followed. The following
suggest that the chemical step is rate limiting for the reaction:
E•G + dSN → products over the range of pH (4.4–7.9) and metal ion
concentrations investigated; there is a constant thio-effect of
2.1 ± 0.2-fold at the pro-RP oxygen (see text; data not shown); (kc/Km)S
for dSN is at least fivefold slower than for dSOH under all the conditions
tested (Figure 3a and data not shown); (kc/Km)S for dSN is log-linear
with pH and parallels the pH dependence of dSH (10 mM Mg2+, 10 mM
Mn2+ / 0.1 mM Mg2+ and 10 mM Mn2+ / 2 mM Mg2+; data not shown).
The observation that the pH dependence of (kc/Km)S for dSN parallels
that for dSH further suggests that deprotonation of the 2′-NH+
3 group to
cm7209.qxd
04/12/2000
01:18
Page 95
Research Paper Cleavage site 2′′-OH in RNA catalysis Yoshida et al.
give dSNH2 does not affect the rate constant of the reaction:
E•G + S → products. If dSNH2 were substantially less reactive than
+
dSNH+3 , (kc/Km)S for dSN would be expected to level off as the 2′-NH3
group deprotonates above pH 6 [20,26], analogous to the leveling-off of
the rate constant for the reaction: E•rSN + G → products (Figure 2). The
absence of a rate effect from deprotonation of the 2′-group in the reaction: E•G + S → products, despite the faster rate of the chemical step
+
from the E•S complex for substrates with 2-NH3 than with 2′-NH2 at
U(–1) (Figure 2), is presumably due to the weaker binding of substrates
+
with 2-NH3 than 2′-NH2 at U(–1). Compared to 2′-OH or 2′-NH2, the 2′+
NH3 group weakens the stability of the P1 duplex formed between S
and the internal guide sequence of E by at least 30-fold ([20,28] and
S.S. & D.H., unpublished results). Thus, the reaction of dSNH2 can be
followed at pH 7.5, without significant contribution from reaction of the
small fraction of dSNH+3 present at this pH (0.05; pKa ~6.2 [20]).
Determination of the pKa of the 2′-NH3+ group in the E•rSN
complex
The different reactivities of rSNH+3 in the reaction: E•S + G → products
+
[(kc / Km)G] provide a signal for deprotonation of the -NH3 group in the
E•rSN complex (Figure 2). The value of pKaE•rSN, the negative logarithm
+
of the observed equilibrium constant for deprotonation of the -NH3
group in E•rSN, was obtained from the pH dependence of (kc / Km)G for
rSN relative to rSOH, (kc / Km)G,rel. The data were fit to equation 3, which
was derived from equation 2:
rel
G,rel
(k c /Km )G,
obsd = (k c /K m )rSNH
G,rel
×
(k c/K m )rS
+
NH 3
2
×
KEa •rSN
E •rSN
Ka
+ [H+ ]
+
[H ]
+
(3)
KEa •rSN + [H+ ]
(kc / Km)G,rel
obsd is the observed second-order rate constant for reaction of
+ and (k / K ) G,rel
rSN relative to rSOH at a particular pH, and (kc / Km)rSG,Rel
NH3
c
m rSNH2
are the pH-independent rate constants for reaction of rSNH+3 and rSNH2,
respectively, relative to rSOH. The pKa of 5.0 obtained for the 2′-amino
group in E•rSN is lower than the solution pKa of this group of 6.2 [20].
This is presumably due to the ~30-fold weaker affinity of the oligonu+
cleotide substrate with 2′-NH3 than with 2′-NH2 at U(–1) (see previous
section), as illustrated in the thermodynamic cycle of equation 4:
(4)
In equation 4, the affinities of rSNH+3 and rSNH2 are expressed as equilibrium association constants relative to rSOH (Kass,rel) in order to
correct for a small pH effect on substrate affinity that is not specific to
the 2′-amino group (2–3-fold). The thermodynamic relationship in equation 4 predicts that the equilibrium for deprotonation of the 2-NH+3
group (Ka) would be 30-fold more favorable on the ribozyme than in
aqueous solution:
KEa•rSN
N
K rS
a
=
NH2
K rS
ass,rel
rSNH +
K ass,rel3
= 30;
(5)
pKaE•rSN = pKarSN – log30 = 6.2 – 1.47 = 4.7. The predicted pKa of 4.7
for the U(–1) 2′-amino group in the E•rSN complex is the same, within
error, as the value of pKaE•rSN = 5.0 determined experimentally (Figure 2).
Acknowledgements
We are grateful to L. Beigelman and F. Eckstein for the gift of oligonucleotide substrates, G. Narlikar for initial results and intellectual input, and
95
members of the Herschlag lab for comments on the manuscript. This work
was supported by grants from NIH to D.H. and HHMI to J.A.P.
References
1. Cech, T.R. & Herschlag, D. (1996). Group I ribozymes: substrate
recognition, catalytic strategies, and comparative mechanistic analysis.
In Nucleic Acids and Molecular Biology. (Eckstein, F. & Lilley, D.M.J.,
eds), pp. 1-17, Springer-Verlag, Berlin.
2. Herschlag, D., Eckstein, F. & Cech, T.R. (1993). The importance of
being ribose at the cleavage site in the Tetrahymena ribozyme
reaction. Biochemistry 32, 8312-8321.
3. Sugimoto, N., Tomka, M., Kierzek, R., Bevilacqua, R.C. & Turner, D.H.
(1989). Effects of substrate structure on the kinetics of circle opening
reactions of the self-splicing intervening sequence from Tetrahymena
thermophila: evidence for substrate and Mg2+ binding interactions.
Nucleic Acids Res. 17, 355-371.
4. Knitt, D.S., Narlikar, G.J. & Herschlag, D. (1994). Dissection of the role
of the conserved G•U pair in group I RNA self-splicing. Biochemistry
33, 13864-13879.
5. Strobel, S.A. & Cech, T.R. (1996). Exocyclic amine of the conserved
G•U pair at the cleavage site of the Tetrahymena ribozyme contributes
to 5′-splice site selection and transition state stabilization.
Biochemistry 35, 1201-1211.
6. Strobel, S.A. & Ortoleva-Donnelly, L. (1999). A hydrogen-bonding
triad stabilizes the chemical transition state of a group I ribozyme.
Chem. Biol. 6, 153-165.
7. Sjögren, A.-S., Pettersson, E., Sjöberg, B.-M. & Strömberg, R. (1997).
Metal ion interaction with cosubstrate in self-splicing of group I
introns. Nucleic Acids Res. 25, 648-653.
8. Sugimoto, N., Kierzek, R. & Turner, D.H. (1988). Kinetics for reaction of a
circularized intervening sequence with CU, UCU, CUCU and CUCUCU:
mechanistic implications from the dependence on temperature and on
oligomer and Mg2+ concentrations. Biochemistry 27, 6384-6392.
9. Smith, D. & Pace, N.R. (1993). Multiple magnesium ions in the
ribonuclease P reaction mechanism. Biochemistry 32, 5273-5281.
10. Shan, S. & Herschlag, D. (1999). Probing the role of metal ions in
RNA catalysis: kinetic and thermodynamic characterization of a metal
ion interaction with the 2′-moiety of the guanosine nucleophile in the
Tetrahymena ribozyme reaction. Biochemistry 38, 10958-10975.
11. Piccirilli, J.A., Vyle, J.S., Caruthers, M.H. & Cech, T.R. (1993). Metal-ion
catalysis in the Tetrahymena ribozyme reaction. Nature 361, 85-88.
12. Weinstein, L.B., Jones, B.C., Cosstick, R. & Cech, T.R. (1997). A
second catalytic metal ion in a group I ribozyme. Nature 388, 805-808.
13. Shan, S., Yoshida, A., Sun, S., Piccirilli, J. & Herschlag, D. (1999).
Three distinct metal ions at the active site of the Tetrahymena group I
ribozyme. Proc. Natl Acad. Sci. USA 96, 12299-12304.
14. Yoshida, A., Sun, S. & Piccirilli, J. A. (1999). A new metal ion
interaction in the Tetrahymena ribozyme reaction revealed by double
sulfur substitution. Nat. Struct. Biol. 6, 318-321.
15. Benkovic, S.J. & Schray, K.J. (1973). Chemical basis of biological
phosphoryl transfer. In The Enzymes. (Boyer, P.P., ed.), pp. 201-238,
Academic Press, New York.
16. Cooperman, B.S. (1976). The role of divalent metal ions in phosphoryl
and nucleotidyl transfer. In Metal Ions in Biological Systems. (Sigel,
H., ed.), pp. 79-125, Marcel Dekker, New York.
17. Admiraal, S. & Herschlag, D. (1995). Mapping the transition-state
for ATP hydrolysis: Implications for enzymatic catalysis. Chem. Biol.
2, 729-739.
18. Holbrook, S. R., Cheong, C., Tinoco, I. & Kim, S.-H. (1991). Crystal
structure of an RNA double helix incorporating a track of non-watsoncrick base pairs. Nature 353, 579-581.
19. Mueller, U., Schubel, H., Sprinzl, M. & Heinemann, U. (1999). Crystal
structure of acceptor stem of tRNAAla from Escherichia coli shows
unique G•U wobble base pair at 1.16º resolution. RNA 5, 670-677.
20. Aurup, H., Tuschl, T., Benseler, F., Ludwig, J. & Eckstein, F. (1994).
Oligonucleotide duplexes containing 2′-amino-2′-deoxycytidines:
thermal stability and chemical reactivity. Nucleic Acids Res. 22, 20-24.
21. Knitt, D.S. & Herschlag, D. (1996). pH Dependencies of the
Tetrahymena ribozyme reveal an unconventional origin of an apparent
pKa. Biochemistry 35, 1560-1570.
22. Celander, D.W. & Cech, T.R. (1991). Visualizing the higher order
folding of a catalytic RNA molecule. Science 251, 401-407.
23. Latham, J.A. & Cech, T.R. (1989). Defining the inside and outside of a
catalytic RNA molecule. Science 245, 276-282.
24. Wang, J.F. & Cech, T.R (1994). Metal ion dependence of active-site
structure of the Tetrahymena ribozyme revealed by site-specific photocross-linking. J. Am. Chem. Soc. 116, 4178-4182.
cm7209.qxd
04/12/2000
96
01:18
Page 96
Chemistry & Biology 2000, Vol 7 No 2
25. Herschlag, D. & Khosla, M. (1994). Comparison of pH dependencies
of the Tetrahymena ribozyme reactions with RNA 2′-substituted and
phosphorothioate substrates reveals a rate-limiting conformational
step. Biochemistry 33, 5291-5297.
26. Shan, S., Narlikar, G.N. & Herschlag, D. (1999). Protonated
2′-aminoguanosine as a probe of the electrostatic environment of the
active site of the Tetrahymena group I ribozyme. Biochemistry
38, 10976-10988.
27. Herschlag, D., Piccirilli, J.A. & Cech, T.R. (1991). Ribozyme-catalyzed
and nonenzymatic reactions of phosphate diesters: rate effects upon
substitution of sulfur for a nonbridging phosphoryl oxygen atom.
Biochemistry 30, 4844-4854.
28. Narlikar, G.J., Khosla, M., Usman, N. & Herschlag, D. (1997).
Quantitating tertiary binding energies of 2′OH groups on the P1
duplex of the Tetrahymena ribozyme: Intrinsic binding energy in an
RNA enzyme. Biochemistry 36, 2465-2477.
29. Sille’n, L.G. & Martell, A.E. (1964). Stability Constants of Metal-Ligand
Complexes, Special publicaton 17. The Chemical Society, London.
30. Martell, A.E. & Smith, R.M. (1976). Critical Stability Constants.
Plenum, New York.
31. Herschlag, D., Eckstein, F. & Cech, T.R. (1993). Contribution of
2′-hydroxyl groups of the RNA substrate to binding and catalysis by
the Tetrahymena ribozyme. An energetic picture of an active site
composed of RNA. Biochemistry 32, 8299-8311.
32. Schreier, A.A. & Schimmel, P.R. (1974). Interaction of manganese
with fragments, complementary fragment recombinations, and whole
molecules of yeast phenylalanine specific transfer RNA. J. Mol. Biol.
86, 601-620.
33. Ott, G., Arnold, L. & Limmer, S. (1993). Proton NMR studies of
manganese ion binding to tRNA-derived acceptor arm duplexes.
Nucleic Acids Res. 21, 5859-5864.
34. Walter, F., Murchie, A.I.H., Thomson, J.B. & Lilley, D.M.J. (1998).
Structure and activity of the hairpin ribozyme in its natural junction
conformation: Effect of metal ions. Biochemistry 37, 14195-14203.
35. Bass, B.L. & Cech, T.R. (1986). Ribozyme inhibitors: deoxyguanosine
and dideoxyguanosine are competitive inhibitors of self-splicing of the
Tetrahymena ribosomal ribonucleic acid precursor. Biochemistry
25, 4473-4477.
36. Tanner, N.K. & Cech, T.R. (1987). Guanosine binding required for
cyclization of the self-splicing intervening sequence ribonucleic acid
from Tetrahymena thermophila. Biochemistry 26, 3330-3340.
37. Crampton, M.R. (1974). Acidity and hydrogen bonding. In Chemistry
of the Thiol Group. (Patai, S., ed.), pp 379-415, John Wiley & Sons,
New York.
38. Abboud, J.-L.M. & Taft, R.W. (1988). Studies on amphiprotic
compounds. 3. Hydrogen-bonding basicity of oxygen and sulfur
compounds. J. Org. Chem. 53, 1545-1550.
39. Meot-Ner, M. & Sieck, L.W. (1986). The ionic hydrogen bond and ion
solvation. 5. OH•••O– bonds. Gas-phase solvation and clustering of
alkoxide and carboxylate anions. J. Am. Chem. Soc. 108, 7525-7529.
40. Sieck. L.W. & Meot-Ner, M. (1989). Ionic hydrogen bond and ion
solvation. 8. RS–•••HOR bond strengths. Correlation with acidities.
J. Phys. Chem. 93, 1586-1588.
41. Allen, F.H., Bird, C.M., Rowland, R.S. & Raithby, P.R. (1997).
Hydrogen-bond acceptor and donor properties of divalent sulfur
(Y-S-Z and R-S-H). Acta Crystallogr. B 53, 696-701.
42. Platts, J.A., Howard, S.T. & Bracke, B.R.F. (1996). Directionality
of hydrogen bonds to sulfur and oxygen. J. Am. Chem. Soc.
118, 2726-2733.
43. Williams, A. (1992). Effective charge and transition-state structure in
solution. Adv. Phys. Org. Chem. 27, 1-55.
44. Narlikar, G.J., Gopalakrishnan, V., McConnel, T.S., Usman, N. &
Herschlag, D. (1995). Use of binding energy by an RNA enzyme for
catalysis by positioning and substrate destabilization. Proc. Natl Acad.
Sci. USA 92, 3668-3672.
45. Christensen, J.J., Rytting, J.H. & Izatt, R.M. (1970). Thermodynamics of
proton dissociation in dilute aqueous solution. Part XV. Proton
dissociation from several monosaccharides at 10 and 40ºC. J. Chem.
Soc. B, 1646-1648.
46. Izatt, R.M., Rytting, J.H., Hansen, L.D. & Christensen, J.J. (1966).
Thermodynamics of proton dissociation in dilute aqueous solution.
V. An entropy titration study of adenosine, pentoses, hexoses, and
related compounds. J. Am. Chem. Soc. 88, 2641-2645.
47. Carter, P. & Wells, J.A. (1988). Dissecting the catalytic triad of a
serine protease. Nature 332, 564-568.
48. Griffin Jr., E.A., Qin, Z., Michels Jr., W.J. & Pyle, A.M. (1995). Group II
intron ribozyme that cleave DNA and RNA linkages with similar
efficiency, and lack contacts with substrate 2′-hydroxyl groups.
Chem. Biol. 2, 761-770.
49. Lambowitz, A.M. & Belfort, M. (1993). Introns as mobile genetic
elements. Annu. Rev. Biochem. 62, 587-622.
50. Yang, J., Mohr, G., Perlman, P.S. & Lambowitz, A.M. (1998). Group II
intron mobility in yeast mitochondria: target DNA-primed reverse
transcription activity of al1 and reverse splicing into DNA transposition
sites in vitro. J. Mol. Biol. 282, 505-523.
51. Matsuura, M., & Lambowitz, A.M. (1997). A bacterial group II intron
encoding reverse transcriptase, maturase, and DNA endonuclease
activities: biochemical demonstration of maturase activity and insertion of
new genetic information within the intron. Genes Dev. 11, 2910-2924.
52. Eickbush, T.H. (1999). Mobile introns: retrohoming by complete
reverse splicing. Curr. Biol. 9, R11-R14.
53. Zaug, A.J., Davila-Aponte, J.A. & Cech, T.R. (1994). Catalysis of RNA
cleavage by a ribozyme derived from the group I intron of Anabaena
pre-tRNALeu. Biochemistry 33, 14935-14947.
54. Hondal, R. J., Bruzik, K. S., Zhao Z. & Tsai, M. D. (1997). Mechanism
of phosphatidylinosito-phospholipase C. 2. Reversal of a thio effect by
site-directed mutagenesis. J. Am. Chem. Soc. 119, 477-478.
55. Zaug, A.J., Grosshans, C.A. & Cech, T.R. (1988). Sequence-specific
endoribonuclease activity of the Tetrahymena ribozyme: enhanced
cleavage of certain oligonucleotide substrates that form mismatched
ribozyme-substrate complexes. Biochemistry 27, 8924-8931.
56. Sun, S., Yoshida, A. & Piccirilli, J. A. (1997). Synthesis of
3′-thioribonucleosides and their incorporation into oligoribonucleotides
via phosphoramidite chemistry. RNA 3, 1352-1363.
57. Herschlag, D. & Cech, T.R. (1990). Catalysis of RNA cleavage by the
Tetrahymena thermophila ribozyme. 1. Kinetic description of the
reaction of an RNA substrate complementary to the active site.
Biochemistry 29, 10159-10171.
58. McConnell, T.S., Cech, T.R. & Herschlag, D.H. (1993). Guanosine
binding to the Tetrahymena ribozyme: thermodynamic coupling with
oligonucleotide binding. Proc. Natl Acad. Sci. USA 90, 8362-8366.
59. Herschlag, D. (1992). Evidence for processivity and two-step binding
of the RNA substrate from studies of J1/2 mutants of the Tetrahymena
ribozyme. Biochemistry 31, 1386-1398.
60. Bevilacqua, P.C., Kierzek, R., Johnson, K.A. & Turner, D.H. (1992).
Dynamics of ribozyme binding of substrate revealed by fluorescencedetected stopped-flow methods. Science 258, 1355-1358.
61. Szewczak, A.A., Ortoleva-Donnelly, L., Ryder, S.P., Moncoeur, E.
& Strobel, S.A. (1998). A minor groove RNA triple helix within the
catalytic core of a group I intron. Nat. Struct. Biol. 5, 1037-1042.
Because Chemistry & Biology operates a ‘Continuous
Publication System’ for Research Papers, this paper has been
published via the internet before being printed. The paper can
be accessed from http://biomednet.com/cbiology/cmb — for
further information, see the explanation on the contents pages.